Making Research Software FAIR with CodeMeta [Part I]

Research software metadata is often fragmented across formats, making it difficult to discover, cite, and reuse software. CodeMeta addresses this challenge by providing a standardized, machine-readable vocabulary (based on JSON-LD and schema.org) that acts as a “metadata exchange layer” across software repositories and tools. This interactive workshop introduces participants to CodeMeta and demonstrates how it can improve the FAIRness of the Research Software. Attendees will: Gain an understanding of the CodeMeta schema and its core concepts Learn how to automatically generate codemeta.json files from existing software repositories using tools such as Github actions, autocodemeta tool, somef, etc… Improve your research software metadata by avoiding common pitfalls Use methods for mapping metadata from other schemas (e.g., package managers, repository metadata) into CodeMeta using structured crosswalks and emerging practices such as SSSOM Discuss approaches to use large language models (LLMs) to generate mappings? Through hands-on exercises, participants will generate CodeMeta for their own software and experiment with schema mappings. The session targets researchers, RSEs, and developers interested in improving software metadata quality and integration across infrastructures such as Zenodo and Software Heritage.